CMOS image sensors are well-known and are being implemented for a long time now. A good summary of CMOS sensors can be found in S. Mendis et al., “Progress in CMOS Active Pixel Sensors”, Proc. SPIE vol.2172, p. 19 (1994) and in E. Fossum, “Active Pixel Sensors: Are CCD's Dinosaurs?”, Proc. SPIE vol. 1900, p. 2 (1994). The latest document explains that it is difficult to achieve large array sizes with charge-coupled devices because of the high charge transfer efficiency required and the high vulnerability to single point defects.
Several CMOS implementations have been explored. Passive pixels have been used in the past because of the smaller dimensions of the pixel cell, but these are inherently noisy. Since the last ten years, active pixels have been the first choice for CMOS pixel designs. An active pixel has amplification means inside the pixel, which offers low noise. A conventional implementation of an active pixel is a three-transistor active pixel design as shown in FIG. 1. The input signal to an active pixel 2 is the radiation intensity at the location of that pixel. The radiation may be any suitable radiation such as optical light, IR light, UV light, high energy particles, X-rays, etc. In the following, embodiments of the present invention will be described with reference to incident light. The incident light intensity is transduced by a photosensitive element such as a photodiode 4 and its associated circuits to an analog voltage at the output line 6 of the pixel 2. The sensing is done via a sensor circuit 8, comprising a reverse-biased photodiode 4 and a reset transistor M1. The photodiode 4 is reset periodically to a fixed bias by means of reset transistor M1, which is coupled between the reverse-biased photodiode 4 and a (positive) power supply VDD. Transistor M1 pre-charges the junction capacitance of the photodiode 4 at the beginning of every integration period when a reset signal reset is applied to the gate G1 of the reset transistor M1. The photodiode 4 collects photogenerated charges, e.g. charge carriers such as electrons (a semiconductor silicon substrate exposed to photons results in a release of charge carriers) and discharges in proportion to the integration period and the photocurrent of the photodiode 4. The current that the photons of the light generate in the photodiode 4 is directly related to the incident light. For a linear device the current generated is preferably proportional to the light intensity.
The connection between the reset transistor M1 and the photodiode 4 is the photodiode node 10. In the embodiment of FIG. 1, a signal integrated in the photodiode 4 is present on the photodiode node 10 and can be consequently sensed by a buffer 12, for example comprising buffer transistor M2, and read out in a conventional line-addressing/column readout fashion by the line select transistor M3. The combination of transistors M2 and M3 is only one possible implementation of a buffer/multiplexer. Many other schemes are possible for that part, and are known to a person skilled in the art.
The column output line 6 may end in a current load or a resistive load (not represented in FIG. 1) and will forward the pixel signal to a column amplifier or another type of amplifier (whereby the type is considered not to be a limitation on the present invention).
A plurality of pixels are arranged in an array to form an imaging device, such as a camera for example. Every semiconductor pixel array has a certain yield. For CMOS active pixel arrays, the yield can be about 80% that may certainly be subject to improvement.